Inverse relationship of Rho kinase and myosin-light chain kinase expression in the aging human detrusor smooth muscle

Background Rho kinase (ROCK) and myosin-light chain kinase (MLCK) are key enzymes in smooth muscle contraction. Previous data have suggested that ROCK contribution to human detrusor contraction is increasing with age. Here, we have analyzed the transcriptional expression of Rho kinase isoforms (ROCK1 and ROCK2) as well as MLCK in the aging human detrusor smooth muscle obtained from resected tissue. Methods Small pieces of macroscopically healthy human detrusor smooth muscle (urothelium-free) were prepared for quantitative real-time reverse transcriptase polymerase chain reaction (RT-PCR). Transcript expression (mRNA level) of the target genes ROCK1, ROCK2 and MLCK was normalized to three common reference genes (glyceraldehyde-3-phosphate dehydrogenase, β-actin, phosphoglycerate kinase 1). Results We found that across all ages the expression level of ROCK (i.e. ROCK1 and ROCK2 together) was almost equal to that of MLCK in the human bladder. Further, ROCK2 showed a significantly higher expression level than ROCK1. Among all subjects, there was no significant correlation of any single target gene to age, but expression levels of ROCK and MLCK were inversely correlated. Moreover, the within-subject analysis revealed that the ROCK-to-MLCK ratio showed a significantly negative correlation to age. Thus, within a given subject, there is a relative ROCK down-regulation and concomitant MLCK up-regulation. Conclusions Together with previous data in human detrusor specimens showing increased ROCK contribution to detrusor contraction, we speculate that the drop of the ROCK-to-MLCK ratio may occur as an attempt to compensate for the increased Rho kinase activity.


Background
Age-related changes in urinary bladder contractility are believed to contribute to dysfunctional micturition such as overactive bladder syndrome (OAB). This is a common urological condition associated with urgency, with or without urge incontinence, and increased micturition frequency and nocturia [1]. OAB prevalence increases with age [2], and in order to find new therapeutic strategies for the management of OAB it is crucial to unravel the pathophysiological mechanisms leading to increased bladder contractility. Hence, it is obvious that there is a need for studies on contraction and relaxation mechanisms in the aging human urinary bladder. During the last decade several groups have focussed on the classical adrenergic and cholinergic mechanisms of detrusor motility. For instance, detrusor relaxation following sympathetic β-adrenoceptor activation was not altered during aging [3], the same was true for cholinergic responses [4]. Although there is no age-related correlation, current medication is still based on antimuscarinics and βsympathomimetics. On the other hand, purinergic receptor-mediated contraction increased with age [4] and may thus be regarded as an attractive candidate mechanism. Beyond receptor activation, intracellular contraction mechanisms could also be accessible targets for pharmacological manipulation. We have recently addressed the key enzymes Rho kinase (ROCK) and myosin-light chain kinase (MLCK) in the aging human bladder and could demonstrate that ROCK contribution to detrusor contractility was positively correlated with age, while MLCK contribution was not altered during aging [5]. Therefore, we hypothesized that transcriptional up-regulation of Rho kinase could be responsible for this age-related effect. To this end, we performed quantitative real-time reverse transcriptase polymerase chain reaction (RT-PCR) in order to determine the mRNA expression levels of the two isoforms of Rho kinase (ROCK1 and ROCK2) and MLCK in detrusor smooth muscle tissue obtained from patients undergoing cystectomy.

Preparation of detrusor smooth muscle strips
Human detrusor samples were obtained from 41 patients with an average age of 67 ± 10 years (mean ± S.D.; range 46-84 years old; 33 male and 8 female patients, Table 1). These specimens (1-2 cm width) were prepared from surgically resected bladder wall obtained from patients who underwent cystectomy. The indication for cystectomy was bladder cancer (urothelial carcinoma) with or without prostate cancer in 35/41 cases, other malignancies with or without infiltration of the urinary bladder in 4/41 cases and neurogenic bladder in 2/41 cases (Table 1). Care was taken to dissect only macroscopically healthy bladder wall tissue for this study, but the anatomical origin could not be standardized due to requirements for routine histopathological examinations especially in radical cystectomy. The specimens always contained urothelium which was removed before small pieces were cut for the PCR analysis. Thus, the tissue used in this study was always obtained from within the detrusor smooth muscle. All in vitro experiments with human material performed in this study were approved by the local ethics committee (University of Rostock), and the informed consent to participate in this study was obtained from each patient.

Quantitative RT-PCR analysis
Small pieces from the detrusor smooth muscle portion of approx. 1-2 mm width were prepared from the human tissue sample and immediately frozen in liquid nitrogen. Care was taken that these pieces for quantitative PCR were urothelium-free. For mRNA isolation, TRIZOL reagent was used, and total RNA was reverse-transcribed using Moloney  Single product amplification was confirmed by melting curve and gel electrophoresis analysis. Then, the efficiency was determined by PCR using serial dilution of the cDNA. Messenger-RNA (mRNA) expression levels were efficiency-corrected and determined by normalizing the target genes (ROCK1, ROCK2, MLCK) with three standard reference genes (GAPDH, ACTB, PGK1), expressed as the mean of 2 -ΔΔCt ± SEM.

Statistics
All data are expressed as means ± SEM. Statistical comparison was performed using the two-tailed Student's ttest (SigmaStat 3.5). The level of significance is indicated by asterisks (*P < 0.05; **P < 0.01).

Results
In the present study, we aimed to investigate the transcriptional expression of Rho kinase (ROCK, isoforms ROCK1 and ROCK2) and myosin-light chain kinase (MLCK) in human detrusor smooth muscle using quantitative RT-PCR, normalized to three commonly used reference genes (glyceraldehyde-3-phosphate dehydrogenase [GAPDH], β-actin [ACTB], phosphoglycerate kinase 1 [PGK1]). As it is shown in Fig. 1a, ROCK2 mRNA levels were significantly higher than ROCK1 mRNA levels (P < 0.01 for all three reference genes, paired two-tailed t-test) indicating that ROCK2 is the predominant Rho kinase isoform in human urinary bladder. The relative mRNA contribution was 15.2 ± 1.9 % for ROCK1, 42.8 ± 3.8 % for ROCK2 and 41.9 ± 4.8 % for  (Fig. 1b). When pooling the relative mRNA quantification for ROCK1 and ROCK2, however, there was no consistent difference between ROCK (i.e. ROCK1 and ROCK2 together) and MLCK expression (P > 0.05 for ACTB and PGK1, P < 0.05 for GAPDH, unpaired two-tailed t-test) suggesting that ROCK and MLCK transcripts were almost equally abundant (Fig. 1b).
Since we have previously found that Rho kinase (ROCK) contribution to cholinergic detrusor contraction increased with age [5], we hypothesized that expression of this enzyme might be up-regulated in the aging detrusor. In order to test this, we performed a linear correlation analysis between relative mRNA levels and age (Fig. 2a,b). However, over the broad range of ages from 46 to 84 years we could not detect a significant correlation between ROCK1 or ROCK2 to any reference gene used (Fig. 2a,b). The Pearson's correlation coefficients for these bivariate analyses are given in Table 3. With respect to the other key enzyme of smooth muscle contraction, the correlation coefficient between MLCK   (Fig. 3a, Table 3), but it is obvious that MLCK expression levels did not consistently show age-dependent changes. One has to take into account that the vast majority of patients included in this study were cystectomized due to bladder cancer. Thus, mRNA expression of the analyzed contraction enzymes could be altered by the adjacent tumor infiltration. We therefore performed a regression analysis between the tumor size (pathological tumor size, pT1-4 in Table 1) and mRNA expression, but did not obtain significant correlation (data not shown). Since 33/41 (80 %) patients were male, we tested whether there was a gender specificity in our data. As also shown in Table 3, the correlation coefficients in this subgroup analysis of male patients were similar to those of the total study population. On the other hand, the subgroup analysis for female patients did not reveal statistically significant correlation coefficients either (data not shown). So far, we have not been able to find robust agedependent differences in mRNA abundance for ROCK1, ROCK2 or MLCK. In these interindividual analyses, however, we have plotted each individual against age and may have missed differential changes of these enzymes  Fig. 1 Expression levels of the contraction enzymes Rho kinase (isoforms ROCK1 and ROCK2) and myosin light-chain kinase (MLCK) in the human detrusor smooth muscle. a Relative mRNA content for target genes ROCK1 (white), ROCK2 (gray) and MLCK (black), expressed as percentage of the reference genes GAPDH, ACTB and PGK1. Note that ROCK2 was significantly more expressed than ROCK1. b Relative mRNA abundance of ROCK1, ROCK2 and MLCK. Note that expression level of ROCK1 and ROCK2 together was similar to that of MLCK within a given subject. We, therefore, calculated the ROCK-to-MLCK ratio of every subject and plotted this ratio against age. As shown in Fig. 3b, this analysis revealed a decrease of the ROCK-to-MLCK ratio during aging. This negative correlation was significant using all three reference genes (Fig. 3b). Thus, we conclude that there is a robust age-dependent drop of the ROCK-to-MLCK ratio in the aging human detrusor smooth muscle. Since ROCK2 was the predominant isoform in human detrusor, we also calculated the correlation coefficients for both isoforms separately (Table 3). These analyses revealed that ROCK2-to-MLCK ratio is consistently inversely related to age, but not the ROCK1-to-MLCK ratio (Table 3). These data also suggest that ROCK2 and MLCK are generally negatively correlated. On average, the correlation coefficient between ROCK2 and MLCK mRNA levels was r = −0.5541 (n = 21, P < 0.01, two-tailed t-test). In summary, across all subjects we found an inverse relationship between ROCK and MLCK transcription, respectively. Within a given subject, however, aging was associated with a relative ROCK down-regulation and concomitant MLCK up-regulation.

Discussion
Smooth muscle contraction involves activation of two key enzymes -Rho kinase (ROCK) and myosin lightchain kinase (MLCK). Both have been demonstrated in human detrusor contraction [5], and the present study was performed in order to quantify the mRNA levels for the both isoforms of Rho kinase (ROCK1 and ROCK2) as well as for MLCK in the aging human detrusor smooth muscle.  Fig. 2 Expression of ROCK1 and ROCK2 is not correlated with age. Relative mRNA content for the target genes ROCK1 (a) and ROCK2 (b), normalized to reference genes GAPDH, ACTB and PGK1 plotted against the patients' age (n = 41). There was no significant correlation between any of these target genes and age Quantified PCR data do necessarily depend on the choice of reference genes used for normalization. In particular, altered expression of a reference gene is a potential pitfall in PCR studies that may substantially contaminate the quantification of the genes of interest. With respect to the present study, selection of reference genes in human bladder tissue primarily derived from bladder cancer patients is not simple [6]. Although we have used macroscopically tumor-free tisse specimens, reference genes could be altered by the adjacent pathology. First, we have used the most commonly used reference gene, glyceraldehyde-3-phosphate dehydrogenase (GAPDH). This, however, may be up-regulated in bladder cancer [7]. Hence, we also used β-actin (ACTB) which was not significantly altered in bladder cancer [6] or prostate cancer [8]. The last reference gene, the glycolysis enzyme phosphoglycerate kinase 1 (PGK1), is a target gene of both the myc oncogene pathway and the hypoxia inducible factor 1α (HIF-1α) and is therefore considered as a marker gene for a number of malignant tissues such as kidney cancer [9] or colon cancer [10]. Recently, HIF-1α activation and subsequent PGK1 upregulation has also been demonstrated in bladder cancer [11]. Taken together, cancer-associated up-regulation of the reference genes used in the present study was (i) definitely present (PGK1), (ii) vaguely present (GAPDH) or (ii) absent (ACTB). As a consequence, altered target gene expression was only accepted to be relevant when it could consistently be observed with all three reference genes.
Obviously, the target genes may also be altered by the malignant pathology present in the vast majority of cases. While care was taken to dissect tissue from the macroscopically healthy bladder wall with non-infiltrated urothelium, we cannot exclude the enzyme expression to be altered in detrusor tissue from bladder cancer patients. At least, the non-selective ROCK inhibitor HA-1077 has been found to be beneficial against urothelial cancer [12] and this finding might be interpreted as an indication for an enhanced expression of this enzyme in bladder cancer. However, we did not observe a significant correlation between tumor size and ROCK1 or ROCK2 mRNA expression.
First, our data demonstrated that transcript levels for ROCK (i.e. ROCK1 and ROCK2 together) were similar to that for MLCK. Hence on the transcriptional level, both key enzymes were equally expressed. This is important to note, since it is consistent with functional data from cholinergic contractions of human detrusor specimens that showed an equal pharmacological effect of ROCK and MLCK inhibition, respectively [5]. Thus, both enzymes are not only equally transcribed, but also equally involved on the functional level. Moreover, the present study has also shown that ROCK2 mRNA is significantly more abundant than ROCK1 mRNA. Hence, the predominant Rho kinase isoform in the human detrusor is ROCK2. This is in contrast to rat detrusor tissue, which showed equal transcript levels for ROCK1 and ROCK2 [13]. The predominance of ROCK2 in the human detrusor is a clinically relevant finding, since Rho kinase inhibitors are currently developed for different potential indications, and research so far has largely been concentrated on the ROCK1 isoformin particular in vascular smooth muscle. The notion that this isoform is implicated in arterial hypertension [14] has launched substantial pharmaceutical interest in developing ROCK1 inhibitors as a new class of antihypertensive medication [15]. Moreover, increased ROCK function was also found to be involved in pulmonary hypertension, and again transcription levels of ROCK1 were more markedly enhanced than transcription levels of ROCK2 in both rat [16] and human [17]. Currently, specific ROCK2 inhibitors are not available, but the present study suggests that ROCK2 might be regarded as potential therapeutic target in overactive bladder syndrome.
Another major finding was that the ROCK-to-MLCK ratio showed a significant negative correlation with aging, even though there was no correlation between any of these genes and age. In other words, within a given subject, there was a relative ROCK downregulation andat the same time -MLCK upregulation. This was clearly an unexpected finding, since ROCK contribution to contraction increased with aging as opposed to the contribution of MLCK [5]. Given the fact that the down-regulation of ROCK in the aged detrusor was primarily due to down-regulation of ROCK2, one can speculate about the mechanism of this agedependent transcriptional alteration. We already know from rodent studies that Rho kinase is involved in bladder hyperactivity [18], but in addition, it was always the ROCK1 isoform that was identified to be up-regulated in rat decompensated detrusor hypertrophy [19] or in altered detrusor contractility of diabetic rabbits [20]. Thus, down-regulation of ROCK2 in aged human detrusor tissue is therefore a rather uncommon example of transcription regulation, and possibly not the primary cause. It is therefore intriguing to speculate whether transcriptional ROCK2 down-regulation might be secondary in order to compensate for an enhanced Rho kinase function in the aging bladder. Further studies are needed to address this question experimentally.

Conclusion
The present study shows that ROCK and MLCK are inversely regulated and that down-regulation of ROCKand in particular ROCK2as occurs during aging is counterbalanced by an up-regulation of MLCK. In conclusion, we suggest that there is a physiological homeostatic regulation of both enzymes which is probably disturbed in the aged detrusor. Therefore, pharmacological intervention such as specific ROCK2 inhibitors might be beneficial to reduce bladder overactivity in the elderly.  Fig. 3 The ROCK-to-MLCK ratio is negatively correlated with age. a Relative mRNA content for the target gene MLCK, normalized to reference genes GAPDH, ACTB and PGK1 plotted against the patients' age (n = 21). There was no consistent correlation between MLCK and age. b Ratio of ROCK expression to MLCK expression, plotted gainst age (n = 21). Note that there was a significant negative correlation between the ROCK-to-MLCK ratio and age using all three reference genes